UHF Source-Tagging at 11M Tags/Year

Why source-tagging at the factory (not the DC)

Tagging in the retailer's DC (distribution centre) was the option this customer used during pilot. It worked but added a $0.04 / unit DC labour cost and a 4-day DC dwell time. Source-tagging — encoding at the apparel supplier's factory in Vietnam / Bangladesh / Indonesia — eliminated both costs and shifted the responsibility for chip presence to the goods receipt audit.

• DC tagging — $0.04 / unit DC labour + 4-day DC dwell. Convenient for pilot, not scalable to 11 M units / year.
• Source tagging at supplier factory — $0.00 incremental DC labour, 0-day DC dwell. Tag is part of the hang-tag artwork SKU.
• Trade-off — retailer needs an EPC encoding standard the suppliers can follow. Customer adopted GS1 SGTIN-96 with the retailer's GS1 prefix.
• Receipt audit — retailer's DC reads a 5% sample of every inbound carton through a portal reader to verify chip presence and EPC correctness.

Why Monza R6-P (and not UCODE 9 / Higgs 9)

Monza R6-P, UCODE 9 and Alien Higgs-9 are all viable for apparel source-tagging. The customer chose Monza R6-P for three reasons:

• Autotune — Monza R6-P automatically detunes the antenna based on the underlying substrate. Apparel substrate varies (jeans denim is wet-looking dielectric, lingerie is air-coupled, jackets have metal zipper interference). Autotune normalises read rate across these substrates without per-SKU antenna design.
• Read-rate consistency — the customer's existing reader fleet (Zebra FX9600 portals + MC9300 handhelds) was already tuned for Monza chips from a prior pilot. Switching chip family would have required re-tuning across 340 stores.
• GS1 compatibility — Monza R6-P supports the GS1 EPC Tag Data Standard 2.0 SGTIN-96 encoding natively, which is the customer's mandated standard.
• Cost — at the customer's annual volume the chip-cost delta vs UCODE 9 was less than $0.002 / unit, immaterial vs the Autotune benefit.

GS1 SGTIN-96 encoding at the supplier factory

Each tag carries a unique GS1 SGTIN-96 EPC = retailer's GS1 company prefix + item reference (SKU) + serial number. The retailer provides each supplier with a serial-number block, the supplier's encoder writes the EPC at the printing-and-attaching step, and a per-carton EPC manifest is uploaded to the retailer's WMS before the carton ships.

• EPC field — GS1 SGTIN-96 = 8-bit header (00110000) + 3-bit filter + 3-bit partition + 24-bit GS1 prefix + 20-bit item reference + 38-bit serial.
• Per-supplier serial-number block — retailer pre-allocates serial-number ranges to each supplier to guarantee uniqueness without supplier-to-supplier coordination.
• Carton manifest — supplier's encoder generates an EPCIS-compatible XML manifest per carton, uploaded to retailer's WMS before despatch.
• DC verification — retailer's DC reads each carton's EPCs at receipt and reconciles against the manifest. Mismatch triggers a chargeback to the supplier.

In-store handheld scans drive the inventory accuracy lift

• 98%Store inventory accuracy after RFID (vs 67% pre-RFID)
• -26%Size-level out-of-stock at full-price SKUs
• -22%Shrink (theft + admin error) YoY chain-level
• 11 MUnits tagged per year across 340 stores

Each store performs a full-floor RFID scan once every 7 days using a Zebra MC9300 handheld. A 100,000-unit store completes the scan in 35–45 minutes, vs the 8 hours a visual cycle count used to take. The inventory delta from the scan is reconciled into the WMS overnight and feeds the size-level replenishment recommendation. The 26% out-of-stock reduction is the operations team's primary KPI for the programme — replenishment recommendations only work if the underlying inventory data is trustworthy.

How the programme was built — inlay, source tagging, rollout

The architecture comes down to three decisions and one discipline: an inlay that reads through denim, lingerie and zipper metal without a per-SKU antenna; an EPC written at the supplier's factory so the tag arrives as part of the hang-tag artwork; and a goods-receipt audit that turns “is the chip actually there” into the supplier's problem to prove. The GS1 EPC encoding guide covers the identifier; the supply-chain solution covers the factory-to-DC leg.

• Chip and tag — a Monza R6-P UHF RAIN inlay (96-bit EPC + 32-bit TID, Autotune-enabled) laminated into an 80 × 50 mm paper hang-tag with CMYK two-sided print, a thermal-transfer EPC barcode and the size/colour SKU.
• Source tagging — the EPC is written at the apparel supplier's factory in Vietnam / Bangladesh / Indonesia, so goods arrive pre-encoded: $0.00 incremental DC labour and a 0-day DC dwell, versus $0.04 / unit and a 4-day dwell when the pilot tagged at the DC.
• Encoding standard — GS1 SGTIN-96 built from the retailer's GS1 company prefix, item reference and serial, with a per-supplier serial-number block so suppliers never coordinate to stay unique.
• Receipt audit — the DC reads a 5% carton sample through a portal reader and reconciles it to the supplier's per-carton EPC manifest; a mismatch triggers a chargeback to the supplier.
• Reader estate — the existing Zebra FX9600 portals and MC9300 handhelds, already tuned for Monza from the pilot, carried the rollout to all 340 stores without a chip-family re-tune.
• Scale — 11 M units / year across women's, men's and children's apparel.

The item-level lifecycle — factory encode to store floor

One tag, five checkpoints. The SGTIN-96 written at the factory is the same identifier the store-floor scan, the endless-aisle order and the returns desk all read — no re-encode, no re-key, no second barcode to reconcile. The inventory-tracking solution documents the store-floor half of this flow.

• Factory encode — the supplier writes the GS1 SGTIN-96 at the print-and-attach step and uploads a per-carton EPC manifest to the retailer's WMS before the carton ships.
• DC verify — the distribution centre reads a 5% carton sample through a portal reader and reconciles it to the manifest; the tag is already on the goods, so DC dwell stays at zero.
• Store receiving — cartons arrive pre-encoded, with no DC tagging step and no $0.04 / unit tagging labour to absorb.
• Floor cycle-count — every store re-scans its full floor every 7 days with an MC9300 handheld: 35–45 minutes for a 100,000-unit store, against the 8 hours a visual count used to take.
• POS and EAS — a returned item's SGTIN flips from ‘sold’ to ‘returned’ at the desk, and the store's HF EAS gates at 8.2 MHz coexist with the UHF RAIN inlays without cross-triggering.

Year-one results — measured against the pre-RFID baseline

The programme had one job: make the inventory number trustworthy. Everything downstream — replenishment, endless-aisle fulfilment, shrink control — is what a number you can trust unlocks, and year one put a figure against each. The wider pattern lives in the retail and apparel industry guide.

• Store inventory accuracy: 67% under pre-RFID visual cycle counts, 98% under weekly handheld scans.
• Size-level out-of-stock at full-price SKUs: -26% — the operations team's primary KPI for the programme.
• Shrink from theft and admin error: -22% year on year at the chain level.
• Markdowns from mis-priced inventory: -18%.
• Endless-aisle fulfilment of online orders from store stock: +14% units.

The deployment at a glance — inlay, tag and measured results

One table for the reviewer who wants the specification and the numbers in the same glance. Every figure is the customer's own reported value — nothing here is modelled or rounded up for the slide.

What transfers — where item-level source tagging pays off

The deciding factors here — a size-and-colour SKU count that defeats visual counting, full-price stockouts that cost real margin, and a supply chain long enough to tag upstream — recur across soft-goods retail. The pattern ports cleanly; compare the chip families first on the UCODE vs Monza vs Higgs page, then map the workflow with the inventory-tracking solution.

• Footwear and accessories — the same size/colour SKU explosion that makes visual counts hopeless and item-level RFID decisive.
• Department and specialty apparel — high full-price out-of-stock sensitivity, where a trustworthy size-level count is the replenishment engine's fuel.
• Any retailer with an upstream supplier network — source tagging moves the encode to the factory and the audit to goods-receipt, exactly as it did here.
• Omnichannel programmes leaning on ship-from-store — endless-aisle fulfilment only works when store stock is accurate enough to promise online.

FAQ